pablodiegoo/Data-Pro-Skill Skills

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

Statistical modeling to isolate true drivers and associations. Use for: (1) Key Driver Analysis, (2) Calculating Chi-Square Residuals, (3) Plotting Association Matrices, (4) Partial Dependence Diagnostics.

Advanced clustering and grouping toolkit using PCA and DBSCAN. Provides a complete pipeline for identifying homogeneous groups in high-dimensional data with built-in quality diagnostics and stability metrics. Use for: (1) Grouping similar entities (assets, products, clients) based on multi-dimensional features, (2) Principal Component Analysis for dimensionality reduction, (3) DBSCAN clustering with noise filtering, (4) Diagnosing clustering pathologies like giant cluster ratio or configuration instability.

Decomposes large Markdown documentation into an optimized .agent context structure. Use this skill when: (1) Starting a new project with a large requirements document, (2) Migrating legacy docs to .agent structure, (3) Refactoring existing context files for better organization, (4) Converting PDFs or long READMEs into agent-friendly files, or (5) Optimizing context window usage by splitting monolithic docs into Tasks, Memories, Workflows, and References.

High-performance data manipulation and transformation using Pandas, Numpy, and DuckDB. Use when Claude needs to: (1) Clean or transform structured data (CSV, Parquet, JSON), (2) Perform large-scale aggregations or analytics, (3) Optimize analysis for performance and memory, (4) Implement the 'Tidy Data' (Wide-to-Long) strategy for reporting.

Generates professional statistical charts (Bar, Pie, Grouped) using Matplotlib and Seaborn. Use this skill to visualize survey data, trends, and distributions for reports.

Document format conversion tool. Import: PDF/DOCX/PPTX → Markdown (with OCR fallback). Export: Markdown → PDF/DOCX (with cover page, themes). Use for: (1) Converting external documents to Markdown, (2) Generating professional PDF/DOCX from Markdown analysis results.

Patterns, templates, academic personas, and Mermaid diagrams for creating world-class technical documentation and reports. Use for: (1) Creating new .md files, (2) Adding Mermaid diagrams, (3) Structuring academic/technical reports, (4) Applying GitHub-style standards, (5) Thesis defense quality control.

Normalization and state-mapping of municipal data and generation of professional choropleth maps of Brazil (UF/State level). Use for: (1) Detecting city/state strings and normalizing names, (2) Attaching regional metadata, (3) Generating professional maps integrating survey data with shapefiles.

Tactical and highly interpretable Machine Learning. Use for: (1) Extracting Feature Importance via Random Forest, (2) Running Permutation Tests, (3) Handling Imbalanced Data (SMOTE).

Use this skill when the user wants to interact with Google NotebookLM (notebooklm.google.com). Triggers on requests to query notebooks, search documents, add sources (URLs, GDrive, PDFs), generate audio podcasts/videos from documents, share notebooks, or manage the NotebookLM library. Use this skill for source-grounded answers without hallucination.

Business-level frameworks and actionable reporting for executives. Use for: (1) Plotting Priority Matrices, (2) Generating Pain Curves, (3) Conversion Funnels, (4) Removing Halo Effects to uncover true sentiment.

Core statistical analysis and pipeline automation for survey datasets. Use for: (1) Running standard Crosstabs, NPS, Top-Box calculations, (2) Generating complete EDA or Analytics notebooks, (3) Quantitative and qualitative processing of questionnaire data.

Comprehensive time-series validation and analysis suite. Handles backtesting of trading and non-trading strategies with support for walk-forward validation (training vs testing windows), performance metric calculation (Sharpe, Drawdown, Win Rate), and event-driven resource allocation simulation. Use for: (1) Validating sequential logic on time-series data, (2) Calculating risk-adjusted performance, (3) Simulating constraints in resource distribution, (4) Detecting look-ahead bias through walk-forward testing.